The average organization spends $55.7 million on SaaS annually and wastes $19.8 million of it. Zylo’s 2026 SaaS Management Index, built on 40 million licenses and $75 billion in managed spend, shows license utilization improved from 47% to 54% over the past year. That means 46% of paid licenses still sit idle.
The improvement is real, but two forces are working against it. Vendor price increases pushed overall SaaS spending up 8% year over year despite flat portfolio sizes. And AI-native SaaS applications grew 108% in spending, creating a second, largely ungoverned procurement channel that flows through employee expense reports rather than IT or procurement.
This guide covers the full lifecycle of SaaS spend management: where the waste comes from, how to see it, how to govern it, and how to negotiate it down.
Table of Contents
- The AI SaaS Problem Nobody Budgeted For
- Four Root Causes of SaaS Waste
- Building Visibility Across Five Data Sources
- 2026 Benchmarks
- SaaS Management Platforms: The 2026 Landscape
- A Three-Tier Governance Model
- The 90-Day Optimization Sprint
- Negotiation in a Consumption-Based World
- The Finance-IT Partnership
The AI SaaS Problem Nobody Budgeted For
AI-native SaaS spending surged 108% overall and 393% inside enterprises with 10,000+ employees. ChatGPT climbed from the #14 most-expensed application in 2023 to #1 in 2025. Expense-based SaaS spending (tools bought on corporate credit cards, outside procurement) jumped 267% year over year.
Three forces are colliding. First, consumption-based pricing became the default: three out of five SaaS companies now use usage-based models, and 46% blend subscriptions with variable consumption charges. Second, vendors added AI features with premium surcharges. Microsoft Copilot carries a 60 to 70% premium over base Microsoft 365 costs, and AI add-ons across platforms increase expenses by 30 to 110% of the base subscription. Third, 78% of IT leaders reported unexpected charges tied to consumption-based or AI features, and 61% were forced to cut planned projects to absorb unplanned SaaS cost increases.
Traditional SaaS audit processes catch unused Salesforce licenses and duplicate project management tools. They rarely detect the $1.2 million average that organizations now spend on AI-native applications, because those purchases bypass procurement entirely.
Four Root Causes of SaaS Waste
Decentralized Procurement Without Central Visibility
Business units now control 81% of SaaS spend while IT directly manages only 15%. Departments with corporate credit cards and self-service SaaS portals procure tools in minutes, and 98% of executives admit to bypassing IT for technology purchases.
The result is a portfolio of 305 applications on average, with large enterprises adding 21 new applications per month. Portfolio counts have stabilized (a 0.07% decline year over year), but spending rose 8% because vendors raised prices and AI features added new cost layers.
Auto-Renewal Traps
Standard contracts require 60 to 90 days’ notice for cancellation with annual price escalators of 3 to 7%. With hundreds of renewals per year, manual tracking is operationally impossible. 79% of IT leaders encountered price increases at renewal in the past 12 months, and 77% experienced unexpected costs that surfaced after a contract was signed.
License Hoarding
Even with usage data showing clear waste, organizations resist reducing licenses. The fear of disrupting productivity keeps unused seats active indefinitely. This behavior locks in the 36% average unused license rate that Zylo tracks across its customer base.
Shadow AI
This is the newest and fastest-growing source of waste. Shadow AI usage is up 156% since 2023, with unsanctioned AI tools sitting on company expense cards for an average of 400 days before detection. 39% of employees use applications not managed by their company on work devices, and AI tools are the fastest-growing category of shadow IT.
Unlike traditional shadow IT (a rogue project management board or an unsanctioned file sharing service), shadow AI generates unpredictable consumption charges. An employee who starts using an AI coding assistant on a free tier can trigger automatic upgrades that multiply across teams without anyone noticing until the quarterly expense reconciliation.
Building Visibility Across Five Data Sources
No single source provides complete visibility into a SaaS portfolio. Mature programs triangulate across five channels:
Financial data from ERP systems, corporate cards, and accounts payable captures actual spending but provides no usage context. This is the starting point, not the answer.
SSO and identity provider logs capture authentication events for 40 to 60% of the portfolio. High-quality data for integrated applications, but SaaS purchased outside SSO (increasingly common with AI tools) stays invisible.
Browser extensions and endpoint agents approach 80 to 90% coverage but raise privacy concerns, particularly in regulated industries. Both Productiv and Zylo offer agent-based discovery for deeper visibility.
API integrations provide detailed usage telemetry for individual applications. Productiv now scans portfolios and contracts automatically to surface AI features and flag policy risks, addressing the AI governance gap that pure financial data misses.
Expense report analysis has become critical in 2026. Zylo’s Index identifies expense-based SaaS as the fastest-growing procurement channel, up 267% year over year. Any visibility strategy that ignores expense data is blind to the largest source of new SaaS spending.
2026 Benchmarks
These figures come from Zylo’s 2026 SaaS Management Index, built on analysis of more than 40 million SaaS licenses and $75 billion in spend under management.
| Metric | Current Average | Best Practice Target |
|---|---|---|
| Annual SaaS spend (median) | $20.6M | Varies by headcount |
| SaaS spend per employee (median) | $9,455 | Below $7,000 |
| Applications per organization | 305 | Below 200 |
| License utilization rate | 54% | Above 80% |
| Annual license waste | $19.8M | Below $5M |
| AI-native SaaS spend | $1.2M avg | Governed and budgeted |
| Renewal savings with SMP | 17% avg | 15 to 25% |
The spend-per-employee figure deserves attention. At $9,455 median, SaaS is one of the largest per-employee cost categories after compensation and real estate. For a 1,000-person company, that translates to $9.4 million in SaaS, of which roughly $4.3 million goes to licenses nobody fully uses.
SaaS Management Platforms: The 2026 Landscape
Gartner published its first Magic Quadrant for SaaS Management Platforms in 2025, signaling that the category has matured past early adoption. Gartner predicts that by 2027, 50% of organizations using multiple SaaS applications will centralize management through an SMP, up from less than 20% in 2021.
Zylo leads in financial intelligence and benchmarking. The 2026 SaaS Management Index (40 million licenses, $75 billion in spend) provides the most comprehensive benchmarking dataset in the market. Strong expense-based SaaS discovery addresses the AI tool expensing gap. Best fit for finance-led optimization programs.
Productiv leads in usage analytics through deep API integrations and recently added AI governance capabilities that scan portfolios for AI features and flag policy risks. Best fit for IT-led programs prioritizing utilization data and AI compliance.
Flexera was named a Leader in the 2026 Gartner Magic Quadrant for SaaS Management Platforms. Its strength is broader IT asset management integration, making it the natural choice for organizations already using Flexera for software license management.
Torii offers accessible pricing with strong workflow automation and fast time to value. A strong mid-market option for organizations building their first SaaS management practice.
BetterCloud focuses on SaaS operations (provisioning, deprovisioning, security workflows) rather than pure spend optimization. Named a Gartner Magic Quadrant Leader in 2025. Best fit for IT operations teams whose primary concern is security and lifecycle management.
CloudEagle combines SaaS management with AI-powered procurement and contract negotiation services, positioning as a full-stack spend management solution.
A Three-Tier Governance Model
Tier 1: Acquisition Controls
Define spending thresholds that trigger formal review ($5,000 to $15,000 is typical for mid-market). Require security assessments and business justification for any new application. Build approval workflows that balance speed with oversight.
The critical 2026 addition: extend acquisition controls to AI tool purchases and consumption-based pricing. A tool with a $20/user/month subscription and an uncapped API usage component can generate costs that bear no relationship to the per-seat price. Governance policies must address both components.
Tier 2: Operational Governance
Require SSO integration for deployments over a defined user threshold. Mandate contract documentation in the SMP. Establish quarterly reporting by application owner. Align with your SaaS management policy and information security frameworks.
Add AI-specific operational policies: which AI tools are approved, what data can flow into them, and what consumption limits apply.
Tier 3: Exit and Optimization
Define utilization thresholds that trigger review (below 40% over 90 days is standard). Establish data extraction requirements and transition plans. Every application needs an accountable business owner answering three questions quarterly: why do we have this, who uses it, and is it delivering value?
Use this matrix when evaluating each application:
| Utilization Level | High Criticality | Low Criticality |
|---|---|---|
| High (above 70%) | Retain; optimize pricing | Verify business case |
| Medium (40 to 70%) | Investigate adoption barriers | Consolidation candidate |
| Low (below 40%) | Mandate adoption or replace | Elimination candidate |
The 90-Day Optimization Sprint
Days 1 to 30 (Discovery): Complete the SaaS inventory using financial, SSO, and expense report data. Reconcile sources and assign owners for the top 50 applications by spend. Calculate baseline utilization. Flag renewals in the next 120 days. Separately inventory AI-native tools purchased through expense reports.
Days 31 to 60 (Analysis): Assess each top application across spend, active licenses, utilization rate, contract terms, renewal date, and owner recommendation. Categorize every application: retain, rightsize, renegotiate, consolidate, or eliminate. For AI tools, determine whether consumption-based pricing is predictable enough to budget or requires usage caps.
Days 61 to 90 (Execution): Execute the plan. Track progress weekly. Start renewal negotiations at least 90 days before contract end. Document realized savings and update baselines.
Organizations executing this methodology typically identify 15 to 25% savings across their top 50 applications. The AI-native category often yields higher savings percentages because it has never been governed at all.
Negotiation in a Consumption-Based World
Traditional SaaS negotiation focused on seat count and per-seat pricing. With 59% of vendors expecting usage-based pricing to grow their revenue share (up from 18% in 2023), the negotiation playbook requires expansion.
Six established levers:
- License elimination of unused seats (typically 15 to 20% in the first optimization cycle)
- Tier rightsizing to appropriate plans (20 to 35% cost reduction per user)
- Application consolidation to eliminate duplicate tools (recovers 40 to 60% of the eliminated app’s cost)
- Contract renegotiation at renewal with competitive benchmarking data (10 to 20% savings)
- Payment term optimization through annual or multi-year commitments (15 to 25% savings)
- Feature reduction to drop unused premium capabilities (25 to 40% savings)
Three new levers for consumption-based contracts:
- Consumption caps with burst provisions set a baseline commitment with agreed overage rates, preventing runaway bills while preserving flexibility
- AI feature opt-out clauses allow you to decline AI add-ons bundled into renewals (Microsoft Copilot’s 60 to 70% premium is the clearest example)
- Usage-based price locks commit to a consumption volume in exchange for fixed per-unit pricing for the contract term
Renewal Timeline
Work backward from each renewal date with standard 60-day notice requirements:
- 90 days out: Generate utilization and consumption reports. Identify the application owner. Review contract terms and research alternatives.
- 75 days out: Complete the owner meeting and document a recommendation.
- 60 days out: Execute the decision. Send termination notice if applicable, or open renegotiation.
- 45 days out: Follow up on pending negotiations and escalate if needed.
- 30 days out: Finalize terms and execute contracts.
- 14 days out: Verify license counts, consumption commitments, and billing details.
Contract Terms Worth Prioritizing
Include these clauses in every SaaS agreement: termination for convenience, mid-contract license adjustment rights, annual price increase caps (3 to 5%), clear usage-based overage terms, and data portability guarantees. For AI-augmented products, add a clause specifying that AI feature pricing is separable from the base subscription.
The Finance-IT Partnership
SaaS spend management fails when Finance and IT operate in isolation. Business units control 81% of SaaS spend, so governance requires a cross-functional model with clear accountability.
| Activity | Finance | IT | Business Unit | Procurement |
|---|---|---|---|---|
| Budget planning | Accountable | Consulted | Responsible | Informed |
| Application discovery | Informed | Accountable | Consulted | Informed |
| Usage monitoring | Consulted | Accountable | Responsible | Informed |
| Renewal decisions | Consulted | Consulted | Accountable | Responsible |
| Contract negotiation | Consulted | Consulted | Consulted | Accountable |
| AI tool governance | Consulted | Accountable | Responsible | Consulted |
The AI tool governance row reflects the 2026 reality: AI cost management has become a distinct discipline requiring its own accountability structure. Organizations that treat AI SaaS as “just another software category” will repeat the SaaS sprawl mistakes of the past decade, compressed into months instead of years.
Organizations that assign clear RACI ownership across all six activities and enforce it through quarterly reviews achieve measurably lower waste rates and higher SaaS ROI than those managing software spend through ad hoc processes.